Build with HeyGen Video.
HeyGen's own video model, run on HeyGen's API. Write a prompt and get a 5 to 15 second clip with its own sound (talking, room tone, effects).
Best for
HeyGen's own video model, run on HeyGen's API. Write a prompt and get a 5 to 15 second clip with its own sound (talking, room tone, effects). Use the playground to set up a single step, then open a workflow when you want to connect more tools.
Inputs explained
- Prompt (required)
- Describe the shot and its sound. Long, specific prompts work best: name the camera, the light, the room tone and any dialogue. In reference mode, refer to what you wired as Picture 1, Video 1 or Audio 1 inside angle brackets. Type: text.
- First Frame (optional)
- Optional. The clip opens on this image and keeps its shape. Cannot be combined with references. Type: image.
- Reference Images (max 9) (optional)
- Products, people or places to keep in a new scene. Up to 9, called Picture 1, Picture 2 and so on in wiring order (inside angle brackets in the prompt). Type: list of image.
- Reference Video 1 (optional)
- Optional clip, called Video 1 in the prompt. HeyGen reads its first 5 seconds, and each wired clip bills 5 seconds at the reference rate. Shape between 1:4 and 4:1. Type: video.
- Reference Video 2 (optional)
- Optional clip, called Video 2 in the prompt. HeyGen reads its first 5 seconds, and each wired clip bills 5 seconds at the reference rate. Shape between 1:4 and 4:1. Type: video.
- Reference Video 3 (optional)
- Optional clip, called Video 3 in the prompt. HeyGen reads its first 5 seconds, and each wired clip bills 5 seconds at the reference rate. Shape between 1:4 and 4:1. Type: video.
- Reference Audio (max 3) (optional)
- Voices or sounds to follow, called Audio 1, Audio 2 and so on. Needs at least one reference image or video next to it. Type: list of audio.
- Duration (seconds) (optional)
- Any whole number of seconds from 5 to 15. Type: number. Minimum: 5. Maximum: 15. Default: 5.
- Resolution (optional)
- Set the resolution used by this model. Type: text. Options: 480p, 768p, 1080p (3x price), 2K (3x price). Default: "768p".
- Aspect Ratio (optional)
- Shape of the clip. Ignored with a First Frame (the clip keeps the image shape). Adaptive copies the first reference and needs a reference wired. 1080p and 2K come in 16:9 or 9:16 only. Type: text. Options: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive. Default: "16:9".
- Prompt Enhancement (optional)
- HeyGen rewrites the prompt before generating. Turn it off when your prompt is already long and exact. Type: text. Options: Turbo (fast), Quality (slower), Off (use my words). Default: "turbo".
- Seed (optional)
- Optional. The same prompt and seed give the same clip, so you can change one thing at a time. Blank picks a random seed. Type: number. Minimum: 0. Maximum: 4294967295.
Tips before you run
Start with the default settings and change one input at a time. Keep the result you like on your board so the next step can use it.
If you add a reference file, use a link the workflow can read. Check the inputs and credit estimate on the board before starting a run.
Working with this model
One node, three modes, picked by what you wire
There is no mode dropdown. What you connect decides it:
| What you wire | What HeyGen does |
|---|---|
| Only a prompt | Text to video: a new scene with its own sound |
| First Frame | Image to video: the clip opens exactly on your image and keeps its shape |
| Any Reference port | Reference to video: a new scene that keeps the product, person or place you gave it |
A First Frame and references in the same run is refused before you are charged. Pick one.
Every clip comes back as an MP4 at 24 fps with sound generated from the prompt: talking, room tone and effects. There is no separate voice step.
What it costs
Billed per second of video, and the number on the node is the number you pay. The ~5 credits per second in Pricing above is the default (768p, text or image to video) rounded: exactly 4.5 credits per second, so a 5 second clip is 23 credits.
| Resolution | Text or image to video | Reference to video |
|---|---|---|
| 480p | 3 credits per second | 6 credits per second |
| 768p (default) | 4.5 credits per second | 9 credits per second |
| 1080p | 13.5 credits per second | 27 credits per second |
| 2K | 13.5 credits per second | 27 credits per second |
Each wired Reference Video also bills 5 seconds at the reference rate, whatever its length. We checked this against HeyGen's bill on a real run: a 10.5 second reference clip was billed as 5 seconds. Reference images and audio cost nothing extra.
Some examples: 5 seconds of text to video at 480p is 15 credits. At 768p it is 23. A 5 second reference run at 480p with one reference clip bills 10 seconds at 6 credits, so 60.
References and the prompt
In reference mode you point at what you wired by its label in the prompt: the first reference image is <Picture 1>, the second <Picture 2>, the first clip <Video 1>, the first audio <Audio 1>. Images, videos and audio are numbered separately, in wiring order.
- Up to 9 images, 3 videos and 3 audio files, and 12 in total.
- Each Reference Video port takes one clip. HeyGen uses only its first 5 seconds, so trim it to the 5 seconds you want first. Clips must be between 1:4 and 4:1 in shape.
- Reference audio needs at least one image or clip next to it.
- Files have to be public https links. Uploads and the output of other nodes already are.
Shape and resolution
Aspect Ratio defaults to 16:9. Adaptive copies the shape of the first reference, so it only works in reference mode, and only at 480p or 768p. At 1080p and 2K the clip is 16:9 or 9:16. With a First Frame the setting is ignored: the clip keeps the image's shape, so crop the image first if you want a different one.
Prompts that work
- Long, specific prompts beat short ones. Name the camera and lens, the light, and what is in the room.
- Describe the sound you want, and say "no music" if you do not want a music bed.
- Keep shots simple: one subject, one place, a still or slow camera. Close hand work and long on-screen text are the weak spots.
- Set a Seed to get the same clip back from the same prompt, then change one thing at a time.
- Prompt Enhancement rewrites your prompt before it runs. Turn it off when your prompt is already exact and you want it followed word for word.
Common questions
How do I use HeyGen Video on Wireflow?
Set the inputs above and choose Open in a workflow. Sign in to save your board, then check its credit estimate before you run.
Can I use HeyGen Video through an API?
Yes. Save a workflow with this model, then call the Wireflow run API with its workflow ID. The API tab has examples you can copy.
What changes the credit price?
The price shown uses the model defaults. Your inputs and model settings can change the total. The workflow shows an estimate before you run.
Can I connect this to another model?
Yes. Open it in a workflow and connect its output to another node with a matching input type.